Progressive Growing of GANs for Improved Quality, Stability, and Variation

PGGAN: 渐进生成高质量、多样性的图像.

Semi-Supervised Learning with Context-Conditional Generative Adversarial Networks

通过上下文条件生成对抗网络实现半监督学习.

Context Encoders: Feature Learning by Inpainting

上下文编码器：通过修补进行特征学习.

Semantic Image Synthesis with Spatially-Adaptive Normalization

通过空间自适应归一化进行语义图像合成.

Reusing Discriminators for Encoding: Towards Unsupervised Image-to-Image Translation

NICE-GAN: 把判别器重用为编码器的图像翻译模型.

最优传输(Optimal Transport)问题与Wasserstein距离

Wasserstein Distance.